Overview
Tenstorrent AI accelerator card with RISC-V based Tensix architecture. Features open-source software stack for AI development without vendor lock-in. Optimized for LLM inference and computer vision workloads.
Key Features
- Enterprise-ready configuration
- Enterprise-ready configuration
- Enterprise-ready configuration
Ideal For
developers adding dedicated AI compute capability to existing systems for inference acceleration.
CPU
x86_64 architecture
Motherboard
PCI Express 4.0 x16 slot, Triple-slot width
Memory
64 GB
Storage
100 GB (≥2TB recommended)
Power Connectors
4+4-pin EPS12V
Total Board Power
Up to 160W
Operating Temperature Range (Die)
0C - 75C
Operating System
Ubuntu version 22.04 (Jammy Jellyfish)
Internet Connection
Required for driver and stack installation
Note
*CPU core count and number of sockets will depend on the amount of host preprocessing and post-processing required before and after the accelerator processing.*
Note
**To check your version, type `cat /etc/os-release`.
$1,099.00
Prices may vary. Verify on vendor site.
Quick Specs
- CPU
- x86_64 architecture
- Motherboard
- PCI Express 4.0 x16 slot, Triple-slot width
- Memory
- 64 GB
- Storage
- 100 GB (≥2TB recommended)
- Power Connectors
- 4+4-pin EPS12V
- Total Board Power
- Up to 160W
Tags
llm-inferencedata-analytics